Meta CEO Mark Zuckerberg reveals Llama 3.3 model

Equipped with 70 billion parameters, new variant of Llama 3.3 model comes on par with performance of its 405-billion-parameter counterpart
An undated image. — iStock
An undated image. — iStock

As the year comes to an end, tech billionaire Mark Zuckerberg took the tech world by storm by dropping the last major iteration of its AI model Meta Llama 3.3.

Equipped with 70 billion parameters, the Meta CEO said the model comes on par with the performance of its 405-billion-parameter counterpart, not to mention the surpassing efficiency.

With the release of the latest, highest-end variant of the Meta Llama 3.3, the company is looking forward to widening its applications by enabling and encouraging developers to improve upon the company’s open-source AI protocols, The Express Tribune noted.

With the tool having been well received by the AI developers community, the tech tycoon reportedly claimed that Llama is the most widely used AI model around the world for it was downloaded more than 650 million times.

The rollout reflects the company's commitment to keeping its AI models open source, eyeing to lay the foundations for progress and the development of greater AI tools in the relevant space.

With such efforts aimed at strengthening its position in the market, Meta aims to establish itself as a formidable contender in other spheres including virtual reality (VR).

Meta's collective efforts with a helping hand from third parties possess the potential to offer its services more explicitly on both AI and VR fronts.